rigid_docking_condensed throws errors with arguments for pocket definition

[adrienchaton]

Hello and thanks for sharing this exciting work,

I am trying rigid docking with OMTRA, however I am blocked wrt. two errors when passing arguments to define the target pocket.

If I try using the --pocket_center argument, when I have a negative x coordinate it fails with

omtra: error: argument --pocket_center: expected one argument

e.g. --pocket_center "1,2,3" or --pocket_center "1,2,-3" do not make issues but --pocket_center "-1,2,3" cannot be parsed correctly.

Alternatively, I try to use a reference ligand to define the pocket, via --pocket_ligand but I get

Error: Aromatic bonds should not be present in the adjacency matrix.

I tried to sanitize and/or kekulize the reference ligand but it does not fix the error.
However, the ligands I want to dock do not trigger this error. I could verify this by running prediction with a wrong --pocket_center that do not contain negative coordinates.

Maybe I am doing something wrong too, would appreciate any hints.
Otherwise I could shift all the coordinates to be positives, but I'd like to clarify if possible.

Thanks!

edit: I am running the latest docker for OMTRA

[adrienchaton]

regarding the argument parser, it seems to be a known issue, I am not sure what can be done about it besides just shifting coordinates of molecules so that no negative x needs to be passed

python/cpython#91158

other repos use separate arguments, e.g. box_center_x ; box_center_y ; box_center_z and I guess this is to alleviate this issue since you don't need the delimiter anymore, or using nargs should also work

[adrienchaton]

I wanted to submit a PR to fix this but it is not permitted, I guess this could be a reasonable fix, and it should automatically handle types and number of coordinates

cli.py

[dkoes]

The restriction on creating pull requests was a mistake - can you try again? All users should now be able to create a PR.

[adrienchaton]

thanks @dkoes I was trying again but I still get
remote: Permission to gnina/OMTRA.git denied to adrienchaton.

I also thought to create a branch/PR through github web, but the "new branch" button doesnt show up

maybe I am doing something wrong.. are you sure the permissions are set?

[dkoes]

I don't see a fork of OMTRA in your github? Normally you fork in github, make your changes, push back to github and the create PR shows up. Is that what you did?

[adrienchaton]

sorry, I was working on my local copy of your repo, since I started with cloning it instead of forking

PR is submitted now : )

[adrienchaton]

btw, I did not check if some example scripts should also be updated accordingly, since the CLI has changed...

let me know if some more actions are needed from my side